Two months ago I started building VidClean alongside a day job. Today it's live on Product Hunt.
The idea was simple: every free video tool I tried either watermarked my files, capped me after a few minutes, or made me create an account just to download my own file. So I built tools that do none of those things. 26 total, covering transcription, noise removal, silence cutting, subtitles, and an AI thumbnail creator that composites your actual face into YouTube-ready thumbnails in 8 seconds.
The real story behind keeping it free: I self-host Parakeet-TDT and Whisper for transcription and DeepFilterNet3 for noise removal on CPU workers for about $50 a month. No GPU. No per-minute API bill. That cost structure is why transcription has no monthly minutes cap. Otter caps free users at 300 minutes because their cost structure forces them to. Mine does not.
What shipped in 2 months:
AI transcription with no monthly cap, no account required
Background noise removal with DeepFilterNet3
Silence removal, subtitles, compression, conversion, and 20 more tools
AI thumbnail creator with face upload. Upload your photo, get a YouTube-ready thumbnail in 8 seconds
Pro tier at $8.99/month for heavier jobs (PROLAUNCH gets you $4.99/month for life until July 30)
Interesting growth signal: ChatGPT became my second largest traffic source before I had any meaningful Google rankings. AI assistants started recommending it before humans did.
2,096+ completions so far. Would love feedback from this community, especially on the Pro tier and what would make you actually pay for a tool like this.
The ChatGPT traffic signal is interesting because it raises a different question than product usage.
The challenge may not be whether people find the tools valuable.
It may be whether the users arriving through AI recommendations are the same users who eventually become paying customers.
Those can be very different validation signals if you're not careful.
Yeah I see your point, however it's at least bringing in organic visitors to see my tools instead of having to spend money on ads.
Definitely.
The free traffic is valuable regardless.
The reason I brought it up is that a lot of products end up treating "people arriving" and "people who were always likely to buy" as the same signal.
Those can look identical early on while leading to very different conclusions about what's actually working.
Happy to share the fuller thought if useful. What's the best email to reach you on?